Progress in FY2011 has been in the following areas: 1. AMYLOID FIBRIL STRUCTURES DERIVED FROM BRAIN TISSUE: We have developed a new protocol for partial purification of amyloid from brain tissue obtained at autopsy, and a new protocol for using this material as a "seed" for growing fibrils from synthetic, isotopically-labeled peptide. With the new protocols, we can create 1 mg fibril samples suitable for solid state NMR and electron microscopy studies, starting with 1 g of brain tissue, in a single fibril growth step. Applying this protocol to fronto-temporal lobe and occipital lobe tissue from a diseased Alzheimer's disease patient, we find that there is a single fibril structure in this tissue, a surprising result. We have performed numerous solid state NMR measurements on these brain-seeded fibrils, with isotopic labeling at specific sites and with uniform isotopic labeling. We now have a full set of structural constraints that will lead to a specific structural model in FY2012. 2. SURPRISING ANTIPARALLEL BETA-SHEET STRUCTURE IN MUTANT BETA-AMYLOID FIBRILS: In collaboration with S.C. Meredith, we have recently shown that the Asp23-to-Asn mutant of human beta-amyloid (D23N mutant, or Iowa mutant) is capable of forming amyloid fibrils that contain antiparallel beta-sheets. This is the first demonstration that a full-length peptide or protein could form fibrils that contain antiparallel (rather than parallel) beta-sheets. In FY2011, we have developed a full molecular structural model for antiparallel D23N-Abeta fibrils, revealing how similar sets of hydrophobic interactions can stabilize either parallel or antiparallel structures. In collaboration with Mattson's group in NIA, we have shown that both parallel and antiparallel structures are neurotoxic in cell cultures. We have shown that antiparallel structures are metastable with respect to eventual conversion to parallel structures.